VDI Matters! (things that deal with VDI)

Hello again,

I came across a customer issue recently where the customer was recommended to install a GPGPU to make Windows Movie Maker (a free video editor in Windows Essentials 2012)  function inside Remote Desktop Services on Server 2008R2.  Unfortunately, Movie Maker still wasn’t working after basic installation of the GPGPU – what’s making this fail?

Initially, the customer was trying to setup Movie Maker in an Educational environment so a class can work with editing and polishing raw video footage.  Being in the Education sector, there isn’t a lot of money to go around, so their solution was cost effective server, with Movie Maker being virtualized in RDS and students accessing the RD sessions with thin clients.

However, due to the requirements of Movie Maker – specifically, requiring DirectX 9c compatibility or newer, it could not be run on stock server hardware.  The integrated video card was a Matrox 200er and even with the Dell drivers installed, the card just doesn’t support DirectX (we tested with this tool:  dxdiag.exe).  Here’s the error message that the customer was getting when trying to run Movie Maker:

When the customer came to Dell ProSupport, he stated that someone had recommended putting in a GPGPU to make it work, but it still wasn’t working.  This was right after school had started for the year and it was becoming an urgent issue.  We actively troubleshot the issue for several hours, just making sure that everything was installed properly, from RDS to Movie Maker to AMD FirePro.  After coming to the end of the troubleshooting path, it was decided to reproduce the issue in our lab and find a workaround/fix for the scenario.

After a bit of scrounging and seeking assistance with scrounging hardware, we came up with an R730 and an AMD FirePro S9150 card.  After doing some beating on it, I finally found the root issue – RemoteFX was needed in the RDS session to be able to offer DirectX capabilities (namely Direct3D Acceleration) to the application.

Following the process to install RemoteFX, which in Server 2008R2 was designed for RD-Virtualization, I got the RemoteFX driver onboard and also had to install the MS Video Capture driver to redirect video into the RemoteFX shim driver for the RD sessions.  Once this was accomplished, console video was also redirected into RemoteFX – an unfortunate side-effect, but you can still administer the server over RDS.

Now, a login through RDS allows Movie Maker to work and we see this in DXDiag:

 

The DDI Version is 11 and we have Direct3D Acceleration capability now, which covers the requirements of Movie Maker (note Movie Maker running in the background).

Success!

Ran across a problem the other day with a customer who was running a “bleeding edge” Microsoft VDI configuration without a cookbook.

They had a nicely spec’d Dell R730 server running Server 2012R2 Datacenter with Hyper-V/RD-Virtualization, 2 nVidia M60 GPGPU cards and Windows 10 Virtual Desktops.  His initial complaint was that when watching YouTube videos on the Virtual Desktops (VD), the video was choppy and stuttery over Internet Explorer, as if the GPGPU cards weren’t working well, but secondarily – Chrome & Firefox were having severe display problems all on their own, rendering those browsers completely useless.

The customer admitted that they didn’t build out the solution from a cookbook or working document, it was just pieced together based on what his Salesperson had sold him.  We sat down and worked through the configuration of the environment.  Everything looked OK:  Hyper-V installed, RD-Virtualization installed, NIC/nVidia drivers up to date, RemoteFX video enabled on the VD’s.

After having a colleague do some issue reproduction work with the M60’s, we found a problem with the re-engineering and design of the nVidia M60 and GRID 2.0 capabilities.  nVidia was very much ahead of the game with their K1 & K2 GPGPU video cards, being the first to market with something like this, working with OpenGL, Cuda & some DirectX, but unfortunately, it would seem that someone in the Marketing department decided that they were leaving money on the table somewhere.

So, with the release of the new GRID 2.0 generation of GPGPU cards, nVidia had separated the hardware and software functions of the GPGPU cards.  In better terms, the card itself with all it’s processing capabilities, can be purchased and dropped into a server/workstation and used for Compute processing, right out of the box.  However, the GRID 2.0 software piece that manages splitting the very large GPGPU card into smaller WDDM video card slices (via vGPU Profiles) will require a per seat license – meaning that you will have to pay extra for every user that will be leveraging a GRID 2.0 GPGPU.  So, your per seat cost for Graphics Enhanced VDI just went up – exactly how much is still yet to be seen, but initial numbers are around $250-300/seat.  Again, this is only for VDI use of the new nVidia cards.

Unfortunately, the customer’s sales rep wasn’t aware of the new GRID 2.0 licensing and the equipment went out the door with the expectation that everything would work like GRID 1.0.

Research into the Chrome and Firefox issue found some additional interesting problems.  What the customer (and us) saw was the presentation of websites with Chrome & Firefox while leveraging Remote FX/GPGPU was: really bad display, images were chopped up and pieces displayed all over the page, frames to infinity (like when you have two opposing mirrors that are pointed at each other) and just an overall unusable experience.  After doing repro of that issue, to make sure it wasn’t the customer’s installation, we were able to reproduce the same issue.  Doing some checking off the FAQ’s from Chrome & Mozilla, we found that there are problems with the latest versions Chrome & Firefox leveraging the RemoteFX video driver.  There are plans to fix this in their base code, but the only workaround is to disable Video Graphics Acceleration in the browsers, which totally defeats the usefulness (read as: expense) of a vGPU.  So, if you are doing some sort of VDI implementation with Remote FX & video acceleration – stick with Internet Explorer or Edge browsers for the time being.

Chrome display issues with Remote FX discussion thread is here.

Several cool announcements from Dell yesterday at Citrix Summit ’16 in the world of VDI Solutions and VDI related products.

• New release of Dell Wyse ThinOS (WTOS), v.8.2 which now supports IM based products much better by leveraging the Citrix HDX RealTim Optimization Pack inside the WTOS image for Lync 2010, 2013 & Skype for Business 2015.  Think of it as a bit of a “merge” of the traditional “full client features” of WTOS with the Citrix “special touches” from the Wyse Xenith OS.  Should be fun to see in action soon!
• Updated version of the Dell Appliance for Wyse – Citrix, which allows you to buy Dell 13G server hardware & deploy a scalable XenDesktop environment via the Dell Quick Start Tool (QST).  This continues Dell’s recent push for simpler deployment of complex IT solutions for the common IT Administrator.

More information here.

I work as a Principle ProSupport Analyst at Dell with specialization in VDI and Virtualization technologies, so I get the opportunity to see & play with some of the coolest technology for the Enterprise.  My admitted bias here is that a lot of the Virtualization products I’ll talk about here are on Dell Platforms – I’ve got to keep my job you know, and I’m pretty sure that neither Dell or my wife will approve of purchasing an HP or IBM/Lenovo server to test stuff out on.  Take it for what it is worth.

I’m currently working through the finer support details of the Dell Precision Appliance for Wyse.  This is a fantastic application of almost Server-Class hardware, High Resolution GPGPU video and Remote Access, all in one package.  The tricky part is that the DPAfW is what is known as a “Reference Architecture” and is kind of a “You need to do 1-8 instances of CAD in your DataCenter and access it remotely?  Well, here’s the parts and a tool to do the basic configuration!  Enjoy!!”

Historically, high-end graphic systems required for CAD rendering, Oil & Gas visualizations, Movie & Game rendering happened on a really nice PC and took a while to do – all while the PC was housed under someone’s desk at work or at home.  The need for data security, protecting the Intellectual Property (IP) of what is being designed, mixed with the need for higher end processing and graphics capability has enabled the migration of these systems into the Datacenter, where the server is protected from physical access, has A/C and conditioned power (with battery backups) and periodic backup of Corporate IP.

Dell has offered a single user version of this for quite a few years in the R5400/R5500/R7610 Precision Rackmount Workstations.  Very high end graphics workstations that can handle the power and heat load of most graphics cards with the advantages of RAID redundancy for storage.  These leveraged Teradici Host Access Cards (PCoIP) cards for remote access and using Dell Optiplex FX100 endpoints – Teradici hardware based thin clients.

The latest version of this is the Dell Precision Appliance for Wyse, which comes in 2 flavors and now leverages DellWyse Teradici based Thin Clients:

• Dedicated GPGPU – R7910 running ESXi 6.0, 3x nVidia Quadro 4200 GPGPU cards, 3x Teradici Host Access Cards, supports up to 3 virtual desktops leveraging Quadro 4200 video and remotely accessed via Wyse P25/P45 endpoints.  Wiring diagram for dual monitor configuration:

 

• Shared GPGPU – R7910 running ESXi 6.0, 2x nVidia K2 GPGPU cards, supports 4 or 8 virtual desktops leveraging 1/2 or 1/4 of each K2 card (via K2 Shared GPGPU Configuration) and accessed via Wyse P25/P45 endpoints.

Both versions of the DPAfW are designed to work with a “Quick Start Tool” that will do basic configuration of ESXi based on what video cards are detected onboard ESXi.  Here’s a nice video walkthrough of the QST by one of it’s developers.

… more to come.

Howdy!  My name is Scott Anderson and I’ve been doing “IT Stuff” for over 25 years now with high level certifications in Microsoft & Citrix technologies.  Well… more like 35+ years of IT if you go back to 1982 when my Dad dropped about $1200 for a RadioShack TRS-80 Model III with 48K of Ram and _dual_ single density floppy drives (I could boot TRS-DOS and run Apps too, without having to swap floppy disks!!).  It was then when I fell in love with computing – managing data – making things go.

From there, I went to Kilgore Jr. College and got my A.S. in Data Processing and moved on to Texas A&M to get a degree in Computer Science.  Sadly, after a semester and a half, we both mutually parted ways on the basis of “not meeting expectations” of each other.  I found I wasn’t cut out to be a “coder” or Electrical Engineer (although it would have been cool), and also found that high level classes that are taught by TA’s or Professors that are focused on research and writing their next awesome book so they can get tenure.  That didn’t make for a very good educational experience.

That’s when my education took a detour through the world of Recreation, Park & Tourism Sciences at Texas A&M.  It’s there that I found how to apply my Analytical skills to the research process, having had the opportunities to work with several gifted Professors on the Research Methodology, Data Collection & Analysis side of the world through the end of my Bachelor’s and on into my Master’s program.  SPSS & SAS were my friends and with the use of DBase IV, I could slice a rather large data set into the appropriate pieces for useful statistical analysis.

After graduating from Texas A&M with my M.S., I headed to the “Silicon Prairie” which is also known as The University of Illinois, to study what else but… Leisure Studies.  Thanks to a Graduate Assistantship working on Departmental computers and servers, I learned that I was more interested in working with end users and helping them use computers instead of the Social Theory of Leisure (no offense, Jack Kelly).

So after dropping out of PhD. school, it was time to get a real job.  After 10.5 years of post-High School education, it was time.  My first real job was doing help desk for CSC-Intelicom in Champaign, IL and my first experience with VDI.  This was in 1996, folks!! CSC had a SCO-Unix server running Hydra (pre-Terminal Services) and end users were accessing Windows 3.11 desktops from mainframe CRT terminals with 10MB Ethernet cards with BootP eproms.  This worked famously, until they got a MS Office Macro virus inside the Hydra virtual system via Outlook, and no one knew what to do – it was virtualized and running on SCO-Unix, so we couldn’t just boot the Hydra system to a floppy and clean it up…  Fortunately, I figured that if the macro virus came in via email, so can the cleanup.  We emailed the virus cleaner to one of the employees using the Hydra environment, detached it from Outlook and ran it inside of the virtual environment.  Problem solved.

I’ve worked several different places since then and I have been working the last 15 years with Dell ProSupport.  Initially, I did Microsoft Platform, MS & Citrix VDI plus VDI EUC support for 12 years, and swapped over to VMWare Hypervisor & Horizon VDI support 3 years ago.  But, I still use the same common sense approach to troubleshooting problems:  lots of questions, define the scope, then get to work fixing it.

My plan with this blog is to share some answers and research that I’ve been doing with Virtual Desktop Infrastructure (VDI) so that someone won’t have to “recreate the wheel” or have to dig in 46 different places to get enough information to do something right the first time.

Thanks for reading,

Scott